Saturday, August 5, 2017: 8:00 AM-11:30 AM
C123, Oregon Convention Center
Organizer:
James B. Grace, U.S. Geological Survey, Wetland and Aquatic Research Center
Speaker:
James B. Grace, U.S. Geological Survey, Wetland and Aquatic Research Center
Structural equation modeling (SEM) is a methodological framework for developing and evaluating network hypotheses about systems. Its emphasis and approach are substantially different from the classical statistical modeling that has served as the cornerstone of much contemporary science, though it builds upon that base. Interest in SEM has surged due to growing awareness of SEM’s capacity to increase our understanding of complex ecological systems. The goals of this introductory workshop are to (a) present the rationale for working with structural equation models, (b) provide an overview of the fundamental principles of SEM, and (c) illustrate basic modeling procedures using the lavaan package in R, as well as using other R packages. Participants will work through a series of guided tutorials that are designed to develop confidence in designing, fitting, and assessing basic models. Participants need basic familiarity with introductory statistics.